On 9 Nov 2009, at 16:00, Ian Tickle wrote:
> When I used binary version 5.5.0109 (compiled by gfortran I assume)
> downloaded from Garib's site as he suggested instead of the 5.5.0102
> version I compiled myself using ifort (because gfortran had given
> trouble in the past), the problem went away. Admittedly I didn't do a
> thorough analysis of the circumstances under which the problem is
> observed - sorry but I was just happy to see it working!
Yes the gfortran compiled version of Refmac 5.5.0106 also fixes it
for me but unfortunately it runs much slower -
Refmac_5.5.0102: End of Refmac_5.5.0102
Times: User: 42.7s System: 1.9s Elapsed: 0:45
Refmac_5.5.0106: End of Refmac_5.5.0106
Times: User: 72.8s System: 53.2s Elapsed: 2:07
I was hoping that because simply changing atom labels in the input
file leads to no restraints being applied to these atoms it might now
be easier to work out where ifort is breaking things!
Huw
--
Dr Huw Jenkins
Astbury Centre for Structural Molecular Biology
Room 8.53 Garstang Building
University of Leeds
Leeds, LS2 9JT
+44(0)113 343 4269
[log in to unmask]
|